US11527301B2 - Method for reading and writing and memory device - Google Patents

Abstract

The embodiments provide a method for reading and writing and a memory device. The method includes: applying a read command to the memory device, the read command pointing to address information; reading data to be read out from a memory cell corresponding to the address information pointed to by the read command; storing the address information pointed to by the read command into a memory bit of a preset memory space if an error occurs in the data to be read out, wherein the preset memory space is provided with a plurality of the memory bits, each of the plurality of memory bits being associated with a spare memory cell; and backing up the address information stored in the preset memory space into a non-volatile memory cell according to a preset rule.

Description

CROSS REFERENCE

This application is a continuation of PCT/CN2020/127978, filed Nov. 11, 2020, which claims priority to Chinese Patent Application No. 202010250063.8, titled “METHOD FOR READING AND WRITING AND MEMORY DEVICE” and filed on Apr. 1, 2020, the entire contents of which are incorporated herein by reference.

TECHNICAL FIELD

The present disclosure relates to the field of semiconductor memory technologies, and more particularly, to a method for reading and writing and a memory device.

BACKGROUND

Semiconductor memories are memory components for storing various data and information. With the increase in circuit complexity, various forms of memory devices are inevitably prone to produce defective or damaged memory cells during manufacturing or during use, resulting in reduced reliability and lifespan of the semiconductor memory devices.

Therefore, how to improve the reliability of the memory devices and prolong the lifespan of the memory devices has become an urgent problem to be solved at present.

SUMMARY

A technical problem to be solved by the present disclosure is to provide a method for reading and writing and a memory device to greatly improve the reliability of the memory device and prolong the lifespan of the memory device.

The present disclosure provides a method for reading and writing. The method includes following steps. A read command is applied to the memory device, wherein the read command points to address information. Data to be read out is read from a memory cell corresponding to the address information pointed to by the read command. The address information pointed to by the read command is stored into a memory bit of a preset memory space if an error occurs in the data to be read out, wherein the preset memory space is provided with a plurality of the memory bits, and each of the plurality of memory bits is associated with a spare memory cell. The address information stored in the preset memory space is backed up into a non-volatile memory cell according to a preset rule.

Further, the preset rule is a preset cycle or a preset action.

Further, the address information pointed to by the read command is not stored in the memory bit of the preset memory space if no error occurs in the data to be read out.

Further, the method for reading and writing also includes: loading the address information backed up in the non-volatile memory cell into the preset memory space after the memory device is enabled.

Further, the step of reading data to be read out from a memory cell corresponding to the address information pointed to by the read command also includes: reading, from the memory cell corresponding to the address information pointed to by the read command, a first error correction code (ECC) corresponding to the data to be read out. A method for determining whether an error occurs in the data to be read out includes: decoding the first ECC to determine whether the error occurs in the data to be read out.

Further, a write command is applied to the memory device, and address information pointed to by the write command is compared with the address information stored in the preset memory space. A write operation is performed on the memory cell corresponding to the address information pointed to by the write command if the address information pointed to by the write command is different from the address information stored in the preset memory space. If the address information pointed to by the write command is identical to the address information stored in the preset memory space, it is stopped performing the write operation on the memory cell corresponding to the address information pointed to by the write command, and a write operation is performed on the spare memory cell corresponding to a first memory bit where the address information is stored in the preset memory space.

Further, the step of performing a write operation on the memory cell corresponding to the address information pointed to by the write command also includes: forming a second ECC corresponding to data to be written in the write operation, and writing the second ECC together with the data to be written into the memory cell corresponding to the address information pointed to by the write command.

Further, after the read command is applied to the memory device, and before the data to be read out is read from the memory cell corresponding to the address information pointed to by the read command, the method also includes: comparing the address information pointed to by the read command with the address information stored in the preset memory space; and performing a read operation on the memory cell corresponding to the address information pointed to by the read command if the address information pointed to by the read command is different from the address information stored in the preset memory space.

Further, if the address information pointed to by the read command is identical to the address information stored in the preset memory space, it is stopped performing the read operation on the memory cell corresponding to the address information pointed to by the read command, and a read operation is performed on the spare memory cell corresponding to a first memory bit where the address information is stored in the preset memory space.

Further, when the error occurs in the data to be read out, if the data to be read out is corrected, the corrected data is stored into the spare memory cell; and if the data to be read out is not corrected, the data to be read out is stored into the spare memory cell.

The present disclosure also provides a memory device, which includes: a command receiving unit, configured to receive a read command or a write command; a memory cell, configured to correspond to address information pointed to by the read command or write command; a spare memory cell; a preset memory space, provided with a plurality of memory bits configured for storing address information corresponding to a memory cell with data error, wherein each of the plurality of memory bits is associated with one spare memory cell; an execution unit, configured to perform a read operation or a write operation on the memory cell or the spare memory cell; and a non-volatile memory cell, configured to back up the address information stored in the preset memory space according to a preset rule.

Further, the memory device also includes an ECC decoding unit, wherein the ECC decoding unit is configured to decode a first ECC corresponding to data to be read out in the read operation and to form a second ECC corresponding to data to be written in the write operation.

Further, the memory device also includes a loading unit, wherein the loading unit is connected to the preset memory space and the non-volatile memory cell, and the loading unit is configured to load the address information backed up in the non-volatile memory cell into the preset memory space.

Further, the memory device also includes a comparison unit, wherein the comparison unit is connected to the command receiving unit and the preset memory space, and the comparison unit is configured to compare the address information pointed to by the read command or write command with the address information stored in the preset memory space.

Further, the execution unit is also connected to the comparison unit, and the execution unit is configured to perform a read operation or a write operation on the memory cell corresponding to the address information pointed to by the read command or write command according to a result outputted by the comparison unit, or to stop performing the read operation or the write operation on the memory cell corresponding to the address information pointed to by the read command or write command, and to perform a read operation or a write operation on the spare memory cell corresponding to a first memory bit where the address information is stored in the preset memory space.

Further, the memory device includes a logic layer and at least one memory layer. The command receiving unit, the preset memory space and the execution unit are arranged in the logic layer. The memory cell is arranged in the memory layer, the spare memory cell is arranged in the memory layer or the logic layer, and the non-volatile memory cell is arranged in the memory layer or the logic layer.

The present disclosure has advantages as below. Address information corresponding to a memory cell with data error is differentiated, in real time, from address information corresponding to a memory cell without data error by using the preset memory space, and the memory cell with data error is replaced by the spare memory cell. In a subsequent read/write operation, based on a fact whether the address information pointed to by the read command or write command is located in the preset memory space, it is selected to perform a read operation or write operation on the memory cell corresponding to the address information or to perform a read operation or write operation on the spare memory cell, which avoids data error or data loss, thereby greatly improving the reliability of the memory device and prolonging the lifespan of the memory device.

In addition, according to the method for reading and writing provided by the present disclosure, the address information stored in the preset memory space is backed up into the non-volatile memory cell according to the preset rule, to serve as a basis for the subsequent read/write operation, which can prevent the address information having been stored in the preset memory space from being re-stored in the preset memory space after the memory device is powered on again. In this way, the running speed of the memory device is greatly increased, it is avoided reassociating an invalid memory cell with one spare memory cell, and the reliability of the memory device is further improved.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG.1 is a schematic flow diagram of a method for reading and writing according to a first embodiment of the present disclosure;

FIG.2 is a schematic flow diagram of a method for reading and writing according to a second embodiment of the present disclosure;

FIG.3 is a schematic flow diagram of a method for reading and writing according to a third embodiment of the present disclosure;

FIG.4 is a schematic structural diagram of a memory device according to a first embodiment of the present disclosure;

FIG.5 is a schematic structural diagram of a memory device according to a second embodiment of the present disclosure;

FIG.6 is a schematic structural diagram of a memory device according to a third embodiment of the present disclosure;

FIG.7 is a schematic structural diagram of a memory device according to a fourth embodiment of the present disclosure;

FIG.8 is a schematic structural diagram of a memory device according to a fifth embodiment of the present disclosure; and

FIG.9 is a schematic structural diagram of a memory device according to a sixth embodiment of the present disclosure.

DETAILED DESCRIPTION

Embodiments of a method for reading and writing and a memory device provided by the present disclosure are described below in detail with reference to the accompanying drawings.

Embodiments of a method for reading and writing and a memory device provided by the present disclosure are described below in detail with reference to the accompanying drawings.

A common method for improving reliability of a memory device includes: encoding data as an error correction code (ECC) before the data is written into the memory device, and simultaneously storing the data and the ECC into the memory device. When reading, both the data and the ECC are read simultaneously, and the ECC is decoded to restore data where an error likely occurs.

However, it is found that the ECC can only correct data when the data is read out, and in a memory there still exists a memory cell with data error. In a subsequent data storage process, if at least one memory cell with data error reoccurs in a memory segment corresponding to the above memory cell with data error, in this memory segment there may exist at least two memory cells with data error. However, the ECC will be unable to correct the error, causing this memory segment to be unserviceable, or even causing the memory device to be unserviceable, thereby having a negative effect on the reliability and lifespan of the memory device.

By study, it is found that when a user uses the memory device, it may be avoided performing a read/write operation on the memory cell with data error (i.e., an invalid memory cell) by differentiating the memory cell with data error in real time and replacing the memory cell with data error by a spare memory cell, which can greatly improve the reliability of the memory device and prolong the lifespan of the memory device. Therefore, the present disclosure provides a method for reading and writing, which not only can differentiate a memory cell with data error in real time, and replace the memory cell with data error by a spare memory cell, but also can back up address information of the memory cell into a non-volatile memory cell according to a preset rule.

In a first embodiment of the method for reading and writing provided by the present disclosure, when performing a read operation, address information corresponding to a memory cell where an error occurs in data to be read out is stored in a preset memory space to differentiate the memory cell where the error occurs in the data to be read out from a memory cell where no error occurs in the data to be read out, and the memory cell with data error is replaced by a spare memory cell. In some embodiments, with reference toFIG.1, a schematic flow diagram of the method for reading and writing according to the first embodiment of the present disclosure is illustrated, and the method for reading and writing includes following steps.

In Step S10, a read command is applied to the memory device, wherein the read command points to address information. For example, the address information pointed to by the read command is denoted by A0.

In Step S11, data to be read out is read from a memory cell corresponding to the address information pointed to by the read command. For example, data to be read out is read from a memory cell corresponding to the address information A0.

In Step S12, it is determined whether an error occurs in the data to be read out.

If the error occurs in the data to be read out, this means that the memory cell is invalid. In this case, the address information pointed to by the read command is stored into a memory bit ADD of apreset memory space10, the preset memory space is provided with a plurality of the memory bits ADD, and each of the plurality of memory bits ADD is associated with one spare memory cell. That is, the address information corresponding to an invalid memory cell is stored into the preset memory space, and the invalid memory cell is replaced by the spare memory cell to improve the reliability of the memory device.

For example, if an error occurs in the data to be read out read from the memory cell corresponding to the address information A0 pointed to by the read command, the address information A0 pointed to by the read command is stored into a memory bit ADD of thepreset memory space10, and the memory cell corresponding to the address information A0 is replaced by one spare memory cell associated with the memory bit ADD. If an error occurs in the data to be read out read from a memory cell corresponding to address information A4 pointed to by the read command, the address information A4 pointed to by the read command is stored into the memory bit ADD of thepreset memory space10, and the memory cell corresponding to the address information A4 is replaced by one spare memory cell associated with the memory bit ADD.

Thepreset memory space10 may be a structure having memory function known to those skilled in the art, such as a static random access memory (SRAM), a dynamic random access memory (DRAM), a magnetic random access memory (MRAM), a register, a latch, and a flip-flop, etc.

Thepreset memory space10 may include a plurality of memory bits ADD, for example, 100 memory bits, 150 memory bits, and 200 memory bits, etc. The number of the plurality of memory bits ADD may be set according to actual demands. In the step of storing the address information pointed to by the read command into a memory bit ADD of the preset memory space, the address information pointed to by the read command may be sequentially or randomly stored into the plurality of memory bits ADD of thepreset memory space10, and each of the plurality of memory bits ADD is associated with one spare memory cell.

For example, in this embodiment, the address information corresponding to the memory cells with data error is denoted by A0, A4, A10, A18, A25, . . . An. In thepreset memory space10, the address information is sequentially stored into a first memory bit, a second memory bit, a third memory bit, and so on. Each of the plurality of memory bits ADD is associated with one spare memory cell. For another example, in another embodiment of the present disclosure, address information corresponding to memory cells with data error is denoted by A0, A4, A10, A18, A25, . . . An. In thepreset memory space10, the address information may be randomly stored into any memory bit of thepreset memory space10. For example, A0 is stored into a fifth memory bit, A4 is stored into the second memory bit, A10 is stored into a sixth memory bit, and so on. Each of the plurality of memory bits ADD is associated with one spare memory cell.

It is to be understood that whether the address information pointed to by the read command is sequentially stored into the plurality of memory bits ADD of thepreset memory space10, or the address information pointed to by the read command is randomly stored into the plurality of memory bits ADD of thepreset memory space10, different memory bits ADD are associated with different spare memory cells, to ensure that the spare memory cells associated with the address information are not repetitive, thereby avoiding data storage error.

Further, there are a variety of methods for associating the first memory bit with one spare memory cell. For example, the first memory bit is directly associated with the spare memory cell through circuit design, or the first memory bit is associated with the spare memory cell through an identification code that can identify the spare memory cell. Those skilled in the art may also use other conventional methods, which are not limited in the present disclosure.

Further, when the error occurs in the data to be read out, if the data to be read out is corrected, the corrected data is stored into the spare memory cell; and if the data to be read out is not corrected, the data to be read out is stored into the spare memory cell. For example, in this embodiment, as shown inFIG.1, the data to be read out is not corrected, and thus the data to be read out is stored into the spare memory cell. In other embodiments of the present disclosure, if the data to be read out is corrected, the corrected data is stored into the spare memory cell.

If no error occurs in the data to be read out, this means that the memory cell is valid. In this case, the address information pointed to by the read command is not stored into the memory bit ADD of thepreset memory space10.

For example, if no error occurs in the data to be read out read from a memory cell corresponding to address information A1 pointed to by a read command, the address information A1 pointed to by the read command is not stored into a memory bit ADD of thepreset memory space10. If no error occurs in the data to be read out read from a memory cell corresponding to address information A2 pointed to by a read command, the address information A2 pointed to by the read command is not stored into a memory bit ADD of thepreset memory space10.

Furthermore, the method for reading and writing provided by the present disclosure also includes a step as below. In Step S14, the address information stored in thepreset memory space10 is backed up into the non-volatile memory cell according to a preset rule. In this embodiment, in the Step S14, the address information stored in a memory bit ADD of thepreset memory space10 is backed up into the non-volatile memory cell according to the preset rule.

The preset rule may be a preset cycle or a preset action.

The preset cycle may be a preset time cycle after the memory device is powered on. For example, after the memory device is powered on, the address information stored in thepreset memory space10 and an identification code corresponding to the address information are backed up into the non-volatile memory cell in a cycle of 10 minutes. That is, after the memory device is powered on, the address information stored in thepreset memory space10 and the identification code corresponding to the address information are backed up into the non-volatile memory cell every ten minutes. Alternatively, the address information stored in thepreset memory space10 and an identification code corresponding to the address information are backed up into the non-volatile memory cell in a cycle of one hour. That is, after the memory device is powered on, the address information stored in thepreset memory space10 and the identification code corresponding to the address information are backed up into the non-volatile memory cell every one hour.

The preset action may be an operation received by the memory device. For example, the preset action may be an operation of shutting down the memory device, an operation of restarting the memory device, and a trigger operation set by a user or system including backing up the address information and marking instructions, etc. The trigger operation may be any operation of issuing a trigger instruction, such as an operation of clicking a certain trigger button by the user, or an operation of disabling a certain trigger button, which is not limited in the present disclosure.

According to the method for reading and writing provided by the present disclosure, address information stored in thepreset memory space10 is backed up into the non-volatile memory cell. If the memory device is powered down, the address information may still be recorded in the non-volatile memory cell and is not erased, such that the address information may be used as a basis for a subsequent read/write operation.

Further, the method for reading and writing provided by the present disclosure also includes following step. In Step S15, the address information backed up in the non-volatile memory cell is loaded into thepreset memory space10 after the memory device is enabled (that is, after the memory device is powered on again). The address information loaded into thepreset memory space10 serves as an initial reference for the subsequent read/write operation, which can prevent the address information having been stored in thepreset memory space10 before the memory device is powered down from being re-stored in thepreset memory space10 after the memory device is powered on again. In this way, it is avoided reassociating an invalid memory cell with one spare memory cell, and thus a running speed of the memory device is greatly increased. In this embodiment, after the memory device is enabled, the address information backed up in the non-volatile memory cell is loaded into thepreset memory space10, and the address information is stored into a memory bit ADD of thepreset memory space10.

Further, the present disclosure also provides a method for determining whether an error occurs in the data to be read out. In some embodiments, the step of reading data to be read out from a memory cell corresponding to the address information pointed to by the read command further includes: reading, from the memory cell corresponding to the address information pointed to by the read command, a first error correction code (ECC) corresponding to the data to be read out. For example, number of bits of data read from the memory cell corresponding to the address information pointed to by the read command is64b+8b, wherein64brepresents the number of bits of the data to be read out, and8brepresents the number of bits of the first ECC. According to a corresponding algorithm, the first ECC is decoded to restore data where an error likely occurs. The algorithm belongs to the existing technologies, and thus is not to be described in detail. Therefore, it may be determined whether an error occurs in the data to be read out by decoding the first ECC.

The present disclosure enumerates a method for determining whether an error occurs in the data to be read out by decoding the first ECC. This method includes following steps. The data to be read out is re-encoded to form a new ECC, and an XOR comparison is performed between the new ECC and the first ECC. If the new ECC is consistent with the first ECC in each bit, this means that no error occurs in the data to be read out, and the memory cell is valid. In this case, the address information corresponding to the memory cell is not stored into the preset memory space. If the new ECC is inconsistent with the first ECC, this means that an error occurs in the data to be read out, and the memory cell is invalid. In this case, the address information corresponding to the memory cell is stored into a memory bit of the preset memory space.

If no error occurs in the data to be read out, the data to be read out is used as output data of the memory device. If an error occurs in the data to be read out, the first ECC may be employed to correct the data to be read out, the corrected data is used as the output data of the memory device, and the corrected data is stored into the corresponding spare memory cell.

According to the method for reading and writing provided by the present disclosure, address information corresponding to a memory cell with data error is differentiated, in real time, from address information corresponding to a memory cell without data error by using thepreset memory space10, and the memory cell with data error is replaced by the spare memory cell. In a subsequent read/write operation, based on a fact whether address information pointed to by the read command or write command is located in thepreset memory space10, it is selected to perform a read operation or write operation on the memory cell corresponding to the address information or to perform a read operation or write operation on the spare memory cell, which avoids data error or data loss, thereby greatly improving the reliability of the memory device and prolonging the lifespan of the memory device.

Further, according to the method for reading and writing provided by the present disclosure, the address information stored in the preset memory space is backed up into the non-volatile memory cell according to the preset rule, and the address information backed up in the non-volatile memory cell is loaded into the preset memory space after the memory device is powered on again. The address information loaded into the preset memory space serves as an initial reference for the subsequent read/write operation, which can prevent the address information having been stored in the preset memory space before the memory device is powered down from being re-stored in the preset memory space after the memory device is powered on again. In this way, it is avoided reassociating an invalid memory cell with one spare memory cell, and thus the running speed of the memory device is greatly increased.

The method for reading and writing provided by the present disclosure also provides a second embodiment. After applying a read command to the memory device, and before reading data to be read out from a memory cell corresponding to the address information pointed to by the read command, the method for reading and writing also includes a comparison step. In some embodiments, with reference toFIG.2, a schematic flow diagram of the method for reading and writing according to the second embodiment of the present disclosure is illustrated.

In Step S20, a read command is applied to the memory device, wherein the read command points to address information.

In Step S21, the address information pointed to by the read command is compared with the address information stored in the preset memory space20. That is, it is determined whether the address information pointed to by the read command is identical to the address information stored in a memory bit of the preset memory space20.

If the address information pointed to by the read command is different from the address information stored in the preset memory space20, this means that the address information pointed to by the read command is not stored in a memory bit ADD of the preset memory space20. In this case, the memory cell corresponding to the address information pointed to by the read command is a valid memory cell, and a read operation is performed on the memory cell corresponding to the address information pointed to by the read command, i.e., Step S22 is performed.

For example, the address information pointed to by the read command is denoted by A1, and the address information A1 pointed to by the read command is compared with the address information stored in the preset memory space20. If the address information A1 pointed to by the read command is different from the address information stored in any memory bit ADD of the preset memory space20, this means that the address information A1 pointed to by the read command is not stored into the memory bit ADD of the preset memory space20. In this case, the read operation is performed on the memory cell corresponding to the address information A1 pointed to by the read command, i.e., Step S22 is performed. In Step S22, the read command points to the address information A1.

If the address information pointed to by the read command is identical to the address information stored in the preset memory space20, this means that the address information pointed to by the read command is stored into the memory bit ADD of the preset memory space20. In this case, it is stopped performing the read operation on the memory cell corresponding to the address information pointed to by the write command, and Step S23 is performed. For example, if the address information A4 pointed to by the read command is identical to the address information stored in a certain memory bit ADD of the preset memory space20, this means that the address information A4 pointed to by the read command is stored in the memory bit ADD of the preset memory space20. In this case, it is stopped performing the read operation on the memory cell corresponding to the address information pointed to by the write command, and Step S23 is performed.

In Step S22, the data to be read out is read from the memory cell corresponding to the address information pointed to by the read command. For example, the data to be read out is read from the memory cell corresponding to the address information A1 pointed to by the read command. This step is the same as Step S11 in the first embodiment. Step S24 is performed after Step S22 is performed.

In Step S23, the read operation is performed on the spare memory cell corresponding to the memory bit ADD where the address information is stored in the preset memory space20. For example, if the address information A4 pointed to by the read command is identical to the address information stored in the preset memory space20, this means that the address information A4 pointed to by the read command is stored in the memory bit ADD of the preset memory space20. In this case, it is stopped performing the read operation on the memory cell corresponding to the address information pointed to by the write command, and the read operation is performed on the spare memory cell corresponding to the memory bit ADD where the address information A4 is stored in thepreset memory space10.

Step S25 is performed after Step S23 is performed.

In Step S24, it is determined whether an error occurs in the data to be read out. This step is the same as Step S12 in the first embodiment. If the error occurs in the data to be read out, this means that the memory cell is invalid. In this case, the address information pointed to by the read command is stored into the memory bit ADD of the preset memory space20. However, if no error occurs in the data to be read out, this means that the memory cell is valid. In this case, the address information pointed to by the read command is not stored into the memory bit ADD of the preset memory space20.

Further, when the error occurs in the data to be read out, if the data to be read out is corrected, the corrected data is stored into the spare memory cell; and if the data to be read out is not corrected, the data to be read out is stored into the spare memory cell. For example, in this embodiment, as shown inFIG.2, the data to be read out is not corrected, and thus the data to be read out is stored into the spare memory cell. In other embodiments of the present disclosure, if the data to be read out is corrected, the corrected data is stored into the spare memory cell. In Step S25, data is outputted. In this step, the outputted data may be data stored in the spare memory cell, and may be corrected or uncorrected data to be read out in the memory cell.

It is to be noted that in another embodiment, after Step S23 is performed, it may be continued to determine whether an error occurs in data read out from a corresponding spare memory cell. The data is outputted if no error occurs in the data read out. If the error occurs in the data read out, a memory location of the address information pointed to by the read command in the preset memory space is changed, i.e., the address information pointed to by the read command is stored in another memory bit, and the data to be read out or the corrected data to be read out may be stored into a spare memory cell associated with the other memory bit.

In Step S26, the address information stored in the preset memory space20 is backed up into the non-volatile memory cell according to a preset rule. This step is the same as Step S14, and the preset rule may be a preset cycle or a preset action.

In Step S27, the address information backed up in the non-volatile memory cell is loaded into the preset memory space20 after the memory device is enabled. This step is the same as Step S15. The address information loaded into the preset memory space20 serves as an initial reference for the subsequent read/write operation, which can prevent the address information having been stored in the preset memory space20 before the memory device is powered down from being re-stored in thepreset memory space10 after the memory device is powered on again. In this way, it is avoided reassociating an invalid memory cell with one spare memory cell, and thus a running speed of the memory device is greatly increased. In this embodiment, the address information backed up in the non-volatile memory cell is loaded into the preset memory space20 after the memory device is enabled, and the address information is stored in a memory bit ADD of the preset memory space20.

In the second embodiment, after a read command is applied to the memory device, address information pointed to by the read command is compared with the address information stored in the preset memory space20 to determine whether the address information pointed to by the read command is located in a memory bit ADD of the preset memory space20, such that the read command may be selectively executed on the memory cell corresponding to the address information pointed to by the read command or the spare memory cell. In this way, the reliability of the memory device can be improved. In addition, after the data to be read out is read out, it may be selected whether to store the address information of the memory cell into a memory bit ADD of the preset memory space20 according to a fact whether an error occurs in the data to be read out, such that address information corresponding to a memory cell with data error is differentiated, in real time, from address information corresponding to a memory cell without data error by using the preset memory space, and the memory cell with data error is replaced by the spare memory cell if an error occurs in the data to be read out. Furthermore, the corrected data is stored into the spare memory cell to provide a basis for a subsequent read/write operation, which further improves the reliability of the memory device and prolongs the lifespan of the memory device.

The method for reading and writing provided by the present disclosure also provides a third embodiment. In the third embodiment, a write operation is performed on the memory device. In some embodiments, with reference toFIG.3, a schematic flow diagram of the method for reading and writing according to the third embodiment of the present disclosure is illustrated.

In Step S30, a write command is applied to the memory device, wherein the write command points to address information. For example, a write command is applied to the memory device, wherein the write command points to address information A0.

In Step S31, the address information pointed to by the write command is compared with the address information stored in thepreset memory space30. That is, it is determined whether the address information pointed to by the write command is stored in a memory bit ADD of thepreset memory space30.

If the address information pointed to by the write command is different from the address information stored in thepreset memory space30, this means that the address information pointed to by the write command is not stored in the memory bit ADD of thepreset memory space30. In this case, a write operation is performed on a memory cell corresponding to the address information pointed to by the write command. If the address information pointed to by the write command is identical to the address information stored in thepreset memory space30, this means that the address information pointed to by the write command is stored in the memory bit ADD of thepreset memory space30. In this case, it is stopped performing the write operation on the memory cell corresponding to the address information pointed to by the write command, and the write operation is performed on a spare memory cell corresponding to the memory bit ADD where the address information is stored in thepreset memory space30.

For example, the address information pointed to by the write command is address information A1, and the address information A1 is compared with the address information stored in thepreset memory space30. If the address information A1 is different from the address information stored in any memory bit ADD of thepreset memory space30, this means that the address information A1 pointed to by the write command is not stored in the memory bit ADD of thepreset memory space30. In this case, a write operation is performed on the memory cell corresponding to the address information A1 pointed to by the write command.

For another example, the address information pointed to by the write command is address information A4, and the address information A4 is compared with the address information stored in thepreset memory space30. If the address information A4 is identical to the address information stored in one memory bit ADD of thepreset memory space30, this means that the address information A4 pointed to by the write command is stored in the memory bit ADD of thepreset memory space30. In this case, it is stopped performing the write operation on the memory cell corresponding to the address information A4 pointed to by the write command, and the write operation is performed on the spare memory cell corresponding to the memory bit ADD where the address information A4 is stored in thepreset memory space30.

In Step S32, the address information stored in thepreset memory space30 is backed up into the non-volatile memory cell according to a preset rule. This step is the same as Step S14, and the preset rule may be a preset cycle or a preset action.

In Step S33, the address information backed up in the non-volatile memory cell is loaded into thepreset memory space30 after the memory device is enabled. This step is the same as Step S15. The address information loaded into thepreset memory space30 serves as an initial reference for the subsequent read/write operation, which can prevent the address information having been stored in the preset memory space20 before the memory device is powered down from being re-stored in thepreset memory space10 after the memory device is powered on again. In this way, it is avoided reassociating an invalid memory cell with one spare memory cell, and thus a running speed of the memory device is greatly increased. In this embodiment, the address information backed up in the non-volatile memory cell is loaded into thepreset memory space30, and the address information is stored into a memory bit ADD of thepreset memory space30.

Further, in this third embodiment, the method for reading and writing also includes: forming a second ECC corresponding to data to be written in the write operation, and writing the second ECC together with the data to be written into the memory cell corresponding to the address information pointed to by the write command. When reading data in the memory cell, both the data and the second ECC are read simultaneously, and the second ECC is decoded to restore data where an error likely occurs.

In this third embodiment, according to the method for reading and writing provided by the present disclosure, after a write command is applied to the memory device, the address information pointed to by the write command is compared with the address information stored in thepreset memory space30 to serve as a basis for performing a write operation on a memory cell or spare memory cell, thereby avoiding performing the write operation on an invalid memory cell. In this way, the reliability of the memory device is improved, and the lifespan of the memory device is prolonged.

The present disclosure also provides a memory device that can implement the above method for reading and writing. The memory device includes, but is not limited to, a volatile memory such as a dynamic random access memory (DRAM) and a static random access memory (SRAM), and a non-volatile memory such as an NAND flash memory, an NOR flash memory, a ferroelectric random access memory (FeRAM), a resistance random access memory (RRAM), a magnetic random access memory (MRAM), and a phase change random access memory (PCRAM), etc.

With reference toFIG.4, a schematic structural diagram of a memory device according to a first embodiment of the present disclosure is illustrated. The memory device includes: acommand receiving unit40, amemory cell41, apreset memory space42, anexecution unit43, aspare memory cell44, and anon-volatile memory cell45.

Thecommand receiving unit40 is configured to receive a read command or write command applied to the memory device, or address information pointed to by the read command or write command.

Thememory cell41 corresponds to the address information pointed to by the read command or write command, and is configured to store data. In the present disclosure, thememory cell41 may be a memory unit well known to those skilled in the art, such as a basic memory cell, a memory segment, memory pages, and memory blocks, which are not limited by the present disclosure.

Thepreset memory space42 is provided with a plurality of memory bits configured for storing address information corresponding to a memory cell with data error, and each of the plurality of memory bits is associated with one spare memory cell. In other embodiments of the present disclosure, thepreset memory space42 is also configured to store an identification code, and the identification code is configured for identifying a spare memory cell. Thepreset memory space42 may be a structure having memory function known to those skilled in the art, such as a static random access memory (SRAM), a dynamic random access memory (DRAM), a magnetic random access memory (MRAM), a register, a latch, and a flip-flop, etc.

Theexecution unit43 is configured to perform a read operation or a write operation on thememory cell41. Theexecution unit43 is also connected to thepreset memory space42, and is configured to selectively perform a read operation or a write operation on thememory cell41 corresponding to the address information or thespare memory cell44 according to a record of thepreset memory space42.

In some embodiments, if the address information pointed to by the read command or write command is different from address information stored in any memory bit of thepreset memory space42, theexecution unit43 performs the read operation or write operation on the memory cell corresponding to the address information. If the address information pointed to by the read command or write command is identical to the address information stored in a memory bit of thepreset memory space42, theexecution unit43 stops performing the read operation or write operation on the memory cell corresponding to the address information, and performs the read operation or write operation on a spare memory cell associated with the memory bit.

Thespare memory cell44 corresponds to the address information stored in the preset memory space, and is configured to store data. In the present disclosure, thespare memory cell44 may be a memory unit well known to those skilled in the art, such as a basic memory cell, a memory segment, memory pages, and memory blocks, which are not limited by the present disclosure.

Thenon-volatile memory cell45 is configured to back up the address information stored in the preset memory space according to a preset rule.

Thenon-volatile memory45 may be a non-volatile memory structure such as NAND, NOR, FeRAM, RRAM, MRAM, and PCRAM, etc.

Thenon-volatile memory cell45 is connected to theexecution unit43, and theexecution unit43 controls the address information stored in thepreset memory space42 to be backed up into thenon-volatile memory cell45 according to a preset rule. The preset rule may be a preset cycle or a preset action. Further, when the preset action is a trigger action inputted by the user, thecommand receiving unit40 can receive the trigger action, and theexecution unit43 can control the address information stored in thepreset memory space42 to be backed up into thenon-volatile memory cell45 according to the trigger action.

Further, theexecution unit43 also can control the address information backed up in thenon-volatile memory cell45 to be loaded into thepreset memory space42. For example, after the memory device is enabled, theexecution unit43 controls the address information (or the address information and an identification code corresponding to the address information) backed up in thenon-volatile memory cell45 to be loaded into thepreset memory space42, to serve as a basis for the subsequent read/write operation, which can prevent the address information having been stored in the preset memory space from being re-stored in the preset memory space after the memory device is powered on again. In this way, it is avoided reassociating an invalid memory cell with one spare memory cell, and the running speed of the memory device is greatly increased.

The memory device of the present disclosure also provides a second embodiment. With reference toFIG.5, a schematic structural diagram of the memory device according to a second embodiment of the present disclosure is illustrated. The second embodiment differs from the first embodiment in that the memory device also includes aloading unit52.

Theloading unit52 is connected to thepreset memory space42 and thenon-volatile memory cell45, and theloading unit52 is configured to load the address information backed up in thenon-volatile memory cell45 into thepreset memory space42. Further, theloading unit52 is also connected to theexecution unit43, and theexecution unit43 controls theloading unit52 to load the address information backed up in thenon-volatile memory cell45 into thepreset memory space42. For example, after the memory device is enabled, theexecution unit43 controls theloading unit52 to load the address information backed up in thenon-volatile memory cell45 into thepreset memory space42. The address information is stored in a memory bit of the preset memory space to serve as a basis for a subsequent read/write operation, which can prevent the address information having been stored in the preset memory space from being re-stored in the preset memory space after the memory device is powered on again. In this way, it is avoided reassociating an invalid memory cell with one spare memory cell, and the running speed of the memory device is greatly increased.

The memory device of the present disclosure also provides a third embodiment. With reference toFIG.6, a schematic structural diagram of the memory device according to a third embodiment of the present disclosure is illustrated. The third embodiment differs from the first embodiment in that the memory device also includes anECC decoding unit46.

TheECC decoding unit46 is connected to theexecution unit43, thememory cell41, and thepreset memory space42.

TheECC decoding unit46 is configured to decode the first ECC corresponding to the data to be read out in the read operation to restore data where an error likely occurs. Furthermore, it may be determined whether to store the address information into a memory bit of the preset memory space based on a fact whether theECC decoding unit46 restores the data. TheECC decoding unit46 is also configured to form a second ECC corresponding to data to be written in the write operation.

In some embodiments, in this embodiment, the command receiving unit receives a write command. When performing the write operation on the memory cell corresponding to the address information pointed to by the write command, theECC decoding unit46 forms an ECC corresponding to data to be written in the write operation, and stores the ECC into thememory cell41 corresponding to the address information pointed to by the write command. After the write operation is completed, when a subsequent read operation is performed on the memory cell corresponding to the address information, theECC decoding unit46 decodes the ECC. It may be determined whether an error occurs in the data to be read out read by the read operation according to decoding of theECC decoding unit46, and then it is determined whether the address information needs to be stored into a memory bit of thepreset memory space42.

Further, in this embodiment, it may be determined whether the address information needs to be stored into the memory bit of thepreset memory space42 based on a fact whether theECC decoding unit46 restores the data. In some embodiments, if theECC decoding unit46 decodes the ECC and restores the data, this means that an error occurs in the data to be read out read by the read operation. In this case, the address information of the memory cell is stored into a memory bit of thepreset memory space42. If theECC decoding unit46 decodes the ECC, but does not restore the data, this means that no error occurs in the data to be read out read by the read operation. In this case, the address information is not stored into the memory bit of thepreset memory space42.

In other embodiments of the present disclosure, theECC decoding unit46 decodes the ECC. Although theECC decoding unit46 does not restore the data, it may be still determined that an error occurs in the data to be read out read by the read operation according to decoding of theECC decoding unit46. In this case, the address information of the memory cell is stored into the memory bit of thepreset memory space42.

In the third embodiment, theexecution unit43 not only is connected to theECC decoding unit46, but also is connected to thememory cell41. However, in a fourth embodiment of the present disclosure, with reference toFIG.7, a schematic structural diagram of the memory device according to the fourth embodiment, theexecution unit43 is connected to theECC decoding unit46, and theECC decoding unit46 is further connected to thememory cell41. It is to be understood that different connection relationships may be selected according to different needs.

The memory device of the present disclosure also provides a fifth embodiment. With reference toFIG.8, a schematic structural diagram of the memory device according to a fifth embodiment of the present disclosure is illustrated. The fifth embodiment differs from the first embodiment in that the memory device also includes acomparison unit49. Thecomparison unit49 is connected to thecommand receiving unit40 and thepreset memory space42, and thecomparison unit49 is configured to compare the address information pointed to by the read command or write command with the address information stored in thepreset memory space42.

Theexecution unit43 is also connected to thecomparison unit49. Theexecution unit43 can perform a read operation or a write operation on the memory cell corresponding to the address information pointed to by the read command or write command according to a result outputted by thecomparison unit49, or theexecution unit43 stops performing the read operation or the write operation on the memory cell corresponding to the address information pointed to by the read command or write command, and performs a read operation or a write operation on the spare memory cell associated with a memory bit where the address information is stored.

For example, thecomparison unit49 compares the address information pointed to by the read command or write command with the address information stored in thepreset memory space42. If the address information pointed to by the read command or write command is different from the address information stored in any memory bit of thepreset memory space42, thecomparison unit49 outputs a signal to theexecution unit43, such that theexecution unit43 performs, according to the signal, the read operation or write operation on the memory cell corresponding to the address information pointed to by the read command or write command. If the address information pointed to by the read command or write command is identical to the address information stored in a memory bit of thepreset memory space42, thecomparison unit49 outputs a signal to theexecution unit43, such that theexecution unit43 stops performing, according to the signal, the read operation or write operation on the memory cell corresponding to the address information pointed to by the read command or write command, and performs the read operation or write operation on the spare memory cell associated with the memory bit.

The memory device of the present disclosure also provides a sixth embodiment. With reference toFIG.9, a schematic structural diagram of the memory device according to the sixth embodiment of the present disclosure is illustrated. The sixth embodiment differs from the third embodiment in that the memory device also includes alogic layer100 and a plurality of memory layers200 (only one memory layer is shown inFIG.9). The plurality ofmemory layers200 may be DRAM chips, and thelogic layer100 may be a layer provided with a logic circuit, such as a control chip or an interposer. The plurality ofmemory layers200 may be vertically stacked above or below thelogic layer100, but the present disclosure is not limited thereto. Furthermore, the plurality ofmemory layers200 may be integrated together using other packaging methods.

Thecommand receiving unit40, thepreset memory space42, theexecution unit43 and theECC decoding unit46 may be all arranged in thelogic layer100. Thememory cell41, thespare memory cell44 and thenon-volatile memory cell45 are arranged in thefirst memory layer200. In another embodiment of the present disclosure, thepreset memory space42 and theECC decoding unit46 may also be arranged in thefirst memory layer200, and thespare memory cell44 and thenon-volatile memory cell45 may be arranged in thelogic layer100.

In this embodiment, thelogic layer100 has at least one firstdata transmission port47, and thefirst memory layer200 has at least one seconddata transmission port48. Instructions and data are transmitted between thelogic layer100 and the plurality ofmemory layers200 through the firstdata transmission port47 and the seconddata transmission port48.

Further, in the sixth embodiment, theexecution unit43 not only is connected to theECC decoding unit46, but also is connected to thememory cell41 through the firstdata transmission port47 and the seconddata transmission port48. However, in a seventh embodiment of the present disclosure, theexecution unit43 is connected to theECC decoding unit46, and theECC decoding unit46 is connected to thememory cell41 through the firstdata transmission port47 and the seconddata transmission port48. It is to be understood that different connection relationships may be selected according to different needs.

It is to be noted that although embodiments of the connection relationships are provided in the above drawings, the present disclosure does not limit the connection manner. It is to be understood that those skilled in the art may select different connection relationships according to different needs.

The memory device of the present disclosure can differentiate address information of an invalid memory cell from address information of a valid memory cell by using the preset memory space when the user uses the memory device, and the invalid memory cell can be replaced by the spare memory cell, such that a read/write operation may be selectively performed on the memory cell or the spare memory cell. In this way, the reliability of the memory device is greatly improved, and the lifespan of the memory device is greatly prolonged.

What is mentioned above merely refers to some embodiments of the present disclosure. It is to be pointed out that to those of ordinary skill in the art, various improvements and embellishments may be made without departing from the principle of the present disclosure, and these improvements and embellishments are also deemed to be within the scope of protection of the present disclosure.

Claims (12)

What is claimed is:

1. A method for reading and writing, comprising:

applying a read command to a memory device, the read command pointing to address information;

reading data to be read out from a memory cell corresponding to the address information pointed to by the read command;

storing the address information pointed to by the read command into a memory bit of a preset memory space if an error occurs in the data to be read out, wherein the preset memory space is provided with a plurality of the memory bits, each of the plurality of memory bits being associated with a spare memory cell;

backing up the address information stored in the preset memory space into a non-volatile memory cell according to a preset rule;

loading the address information backed up in the non-volatile memory cell into the preset memory space after the memory device is enabled;

applying a write command to the memory device, and comparing address information pointed to by the write command with the address information stored in the preset memory space;

performing a write operation on a memory cell corresponding to the address information pointed to by the write command if the address information pointed to by the write command is different from the address information stored in the preset memory space; and

stopping performing the write operation on the memory cell corresponding to the address information pointed to by the write command, and performing a write operation on the spare memory cell corresponding to a first memory bit where the address information is stored in the preset memory space if the address information pointed to by the write command is identical to the address information stored in the preset memory space.

2. The method for reading and writing according toclaim 1, wherein the preset rule is a preset cycle or a preset action.

3. The method for reading and writing according toclaim 1, wherein the address information pointed to by the read command is not stored in the memory bit of the preset memory space if no error occurs in the data to be read out.

4. The method for reading and writing according toclaim 1, wherein the reading data to be read out from a memory cell corresponding to the address information pointed to by the read command further comprises: reading, from the memory cell corresponding to the address information pointed to by the read command, a first error correction code (ECC) corresponding to the data to be read out; and

wherein a method for determining whether an error occurs in the data to be read out comprises: decoding the first ECC to determine whether the error occurs in the data to be read out.

5. The method for reading and writing according toclaim 1, wherein the performing a write operation on the memory cell corresponding to the address information pointed to by the write command further comprises:

forming a second ECC corresponding to data to be written in the write operation, and writing the second ECC together with the data to be written into the memory cell corresponding to the address information pointed to by the write command.

6. The method for reading and writing according toclaim 1, wherein after the applying a read command to the memory device, and before the reading data to be read out from a memory cell corresponding to the address information pointed to by the read command, the method further comprises:

comparing the address information pointed to by the read command with the address information stored in the preset memory space; and

performing a read operation on the memory cell corresponding to the address information pointed to by the read command if the address information pointed to by the read command is different from the address information stored in the preset memory space.

7. The method for reading and writing according toclaim 6, comprising: stopping performing the read operation on the memory cell corresponding to the address information pointed to by the read command, and performing a read operation on the spare memory cell corresponding to a first memory bit where the address information is stored in the preset memory space if the address information pointed to by the read command is identical to the address information stored in the preset memory space.

8. The method for reading and writing according toclaim 1, wherein when the error occurs in the data to be read out, if the data to be read out is corrected, the corrected data is stored into the spare memory cell; and if the data to be read out is not corrected, the data to be read out is stored into the spare memory cell.

9. A memory device, comprising:

a command receiving unit, configured to receive a read command or a write command;

a memory cell, configured to correspond to address information pointed to by the read command or write command;

a spare memory cell;

a preset memory space, provided with a plurality of memory bits configured for storing address information corresponding to a memory cell with data error, each of the plurality of memory bits being associated with one spare memory cell;

an execution unit, configured to perform a read operation or a write operation on the memory cell or the spare memory cell;

a non-volatile memory cell, configured to back up the address information stored in the preset memory space according to a preset rule; and

a comparison unit, wherein the comparison unit is connected to the command receiving unit and the preset memory space, and the comparison unit is configured to compare the address information pointed to by the read command or write command with the address information stored in the preset memory space; and, wherein

the execution unit is further connected to the comparison unit, and the execution unit is configured to:

perform a read operation or a write operation on the memory cell corresponding to the address information pointed to by the read command or write command according to a result outputted by the comparison unit,

stop performing the read operation or the write operation on the memory cell corresponding to the address information pointed to by the read command or write command, and

perform a read operation or a write operation on the spare memory cell corresponding to a first memory bit where the address information is stored in the preset memory space.

10. The memory device according toclaim 9, further comprising an ECC decoding unit, configured to decode a first ECC corresponding to data to be read out in the read operation and to form a second ECC corresponding to data to be written in the write operation.

11. The memory device according toclaim 9, further comprising a loading unit, wherein the loading unit is connected to the preset memory space and the non-volatile memory cell, and the loading unit is configured to load the address information backed up in the non-volatile memory cell into the preset memory space.

12. The memory device according toclaim 9, comprising a logic layer and at least one memory layer, wherein the command receiving unit, the preset memory space and the execution unit are arranged in the logic layer, the memory cell being arranged in the memory layer, the spare memory cell being arranged in the memory layer or the logic layer, and the non-volatile memory cell being arranged in the memory layer or the logic layer.

Images